Instructions to use failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5")
model = AutoModelForCausalLM.from_pretrained("failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5

SGLang

How to use failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5 with Docker Model Runner:
```
docker model run hf.co/failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5
```

Tried Llama 3.1 70B?

by noisefloordev - opened Jul 29, 2024

Discussion

noisefloordev

Jul 29, 2024

I see a bunch of 3.1 8B out there, but nobody's released a 70B yet that I can find. Maybe people are having a harder time with it for some reason. I might make a go at it eventually, but you had better results than I did with 3, so just curious if you've tried (or plan to).

AIGUYCONTENT

Aug 2, 2024

I really like this model (Meta-Llama-3-70B-Instruct-abliterated-v3.5 ) and am hoping someone will "ablate" the 3.1 70B model.

noisefloordev

Aug 5, 2024

I spent a few days trying, but as with Llama 3 I can't reproduce what was done with this model. The best refusal_dir seems to be blocks.29.hook_resid_post, which matches up with this model (only layer 29 was changed). But I can only get any results by editing a lot of layers, if I only apply it to layer 29 it has no noticeable effect. It's probably something I'm doing wrong and not anything that's changed in the model, since I got the same results with 3.0.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment